Supports breaking off

[Scatterplot]

(I posted this question to the Kickstarter comments as well- double posting for exposure since I'm not sure how many people visit the KS comments vs. this board (yet). I'll post up whatever I end up finding out)

I've been trying to work with some models that need supports. I tried auto-adding them in CW, but I've tried two different models now and the supports keep breaking off right as the thin part is getting formed (this is the part that's 0.5 mm by default in CW). (I haven't yet calibrated my printer, so my scaling could be a little off)

Still, it looks like it forms the sort of conic-shaped base OK, but once the really thin beams start to form they just break right off and won't actually form anything other than a tiny dot on top of the base cone area.

I'm using SubSF at 3500ms / 8000ms with 10 base layers. Any suggestions to get supports to work?

[Glacian22]

Some details about your printer would be helpful. What are you using for a vat coating material? Tilt peel or straight lift? What resolution are you printing at? And a screenshot of the layout in CW might help too.

[Scatterplot]

It's the LittleRP. It's straight lift off PDMS. Resolution is... whatever 1080p gets me at the LRP size, not sure what the pixels to DPI number is offhand.

Do you want the config info from CW, the 3D with supports info, or the slice info? I can get a screenshot later tonight when I try again.

(Sorry for the lack of troubleshooting info. The LRP is my first resin printer so I'm still getting the hang of everything. I'm decent with filament printers, just new here, so I'm trying to stay in the LRP subforum for now )

[Praesmeodymium]

sounds like undercure, the default settings of time for the resin are going to vary a little depending on many factors. I would print the calibration test block and see if it makes all 5 of the little sprues, if not then you are underexposing.

You can find the calibration block here

[Scatterplot]

I'll try that tonight, thanks

[Scatterplot]

OK finally got that calibration pic:


The letters look great on top, and I can clearly see four holes. I can't tell if the smallest hole goes all the way through or not, but it looks decent. There are three gaps on the right hand side clearly formed.

The letters on the front are generally OK, but I can see some visible layers, and the smallest column is only partially formed.

The calibration grid also didn't form right for some reason- the back left base pad did form, but it came off the plate when I cleaned it. Three other columns didn't form at all, and only the rightmost three and the back middle (not visible in that pic) actually formed.

I'm guessing that my issue is undercuring, is that correct? Any suggestions to how long to cure for? I'm using 10 base layers at 8000 ms each, then doing 50 micron slices at 3800 ms each. I'm going to try some more cure times, but I'd love some suggestions to keep from having to just guess.

Thanks for the help

[Praesmeodymium]

just gonna have to guess each projector and each bulb are going to be a little different in terms of dose. I was at 8000 for the base layer and I do 5, and 4500 to 5000 for the SF+ red per layer. it does sound like undercure. the grid is tougher print than the little cube, the fact some came off is likely an issue with leveling the build plate. With the big screw loosened push the build plate all the way down to the vat floor and tighten it in place. that should help get it level.

[Scatterplot]

It did actually look quite level. Doesn't the fact that the wider base pads adhered to the build plate (except the one I knocked off during cleaning) mean that the initial leveling was good enough? Maybe I'm not understanding things.

Edit: Whoops looks like I formatted my last post wrong and my image wasn't showing up. Fixed it.

[goopyplastic]

The smallest column should be fully formed (a little floppy is ok and a sign of the minimum acceptable cure time but the column should be the full length and not broken), it does indeed look like undercurring. Undercurring will leave a rough finish also as some of the surface disentegrates during the layer separation stresses.